Reproduce “Physics-Aware Combinatorial Assembly Sequence Planning using Data-free Action Masking” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Ruixuan Liu, Alan Chen, Weiye Zhao, Changliu Liu

The authors present a Lego assembly planner that filters placements before a learned policy selects them. The data-free online mask checks target fit, collision, inventory, insertion feasibility, and modeled structural stability, without a physics engine. In physical human assembly tests, plans succeeded for all 280 shapes across four sets. This result is limited to selected shapes and fixed brick inventory.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.